Effizientes Rendern von Wäldern mittels Gruppierter Detailgrade in 3D-Geoinformationssystemen

Information

• Publication Type: Bachelor Thesis
• Workgroup(s)/Project(s): not specified
• Date: May 2024
• Date (Start): January 2024
• Date (End): May 2024
• Matrikelnummer: 11808244
• First Supervisor: Eduard Gröller

Abstract

Interactive visualization is important for many workflows. Especially so in the context of 3D geo-informations systems, where large quantities of data have to be processed and presented to the user at interactive speeds for productivity and orientation in the geo-spatial context. In heavily forested countries like Austria enormous amounts of geometry have to be drawn when visualizing forests. Naïve rendering approaches fail, even when using heavily simplified geometry for the individual trees. The region in which details are necessary is small and changes frequently. A major part of the scene is far away and needs little detail. These constraints are what this thesis attempts to find a solution for. Thus each tree is represented by a billboard, if not close to the camera. To decrease the computational complexity of selecting the appropriate level of detail for all trees, they are grouped into batches, for which frustum culling and level of detail selection happens. This new approach is implemented, qualitatively evaluated, and compared with existing alternative approaches. Comparison of the approaches on a stress test scene shows that our new approach can be between 1.7 and 6 times faster than the approaches tested against depending on the scenario, while barely reducing visual quality.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@bachelorsthesis{eichner-2024-erv,
  title =      "Effizientes Rendern von W\"{a}ldern mittels Gruppierter
               Detailgrade in 3D-Geoinformationssystemen",
  author =     "Anand Eichner",
  year =       "2024",
  abstract =   "Interactive visualization is important for many workflows.
               Especially so in the context of 3D geo-informations systems,
               where large quantities of data have to be processed and
               presented to the user at interactive speeds for productivity
               and orientation in the geo-spatial context. In heavily
               forested countries like Austria enormous amounts of geometry
               have to be drawn when visualizing forests. Naïve rendering
               approaches fail, even when using heavily simplified geometry
               for the individual trees. The region in which details are
               necessary is small and changes frequently. A major part of
               the scene is far away and needs little detail. These
               constraints are what this thesis attempts to find a solution
               for. Thus each tree is represented by a billboard, if not
               close to the camera. To decrease the computational
               complexity of selecting the appropriate level of detail for
               all trees, they are grouped into batches, for which frustum
               culling and level of detail selection happens. This new
               approach is implemented, qualitatively evaluated, and
               compared with existing alternative approaches. Comparison of
               the approaches on a stress test scene shows that our new
               approach can be between 1.7 and 6 times faster than the
               approaches tested against depending on the scenario, while
               barely reducing visual quality.",
  month =      may,
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Research Unit of Computer Graphics, Institute of Visual
               Computing and Human-Centered Technology, Faculty of
               Informatics, TU Wien ",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2024/eichner-2024-erv/",
}